Two-State Options

John Norstad
j-norstad@northwestern.edu
http://www.norstad.org
January 12, 1999
Updated: November 3, 2011
Abstract
How options are priced when the underlying asset has only two possible future
states. Studying these trivial options helps develop insight into how real options
and the Black-Scholes equation work. We learn about arbitrage, the law of one
price, hedging, risk-aversion, and risk-neutral valuation.
CONTENTS 1
Contents
1 Two-State Options 2
2 The Algebra of Two-State Call Options 9
3 Risk-Neutral Valuation 11
List of Figures
1 Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1 TWO-STATE OPTIONS 2
1 Two-State Options
This note explores some very simple examples of two-state options, where
the underlying asset has only two possible future prices. Our analysis goes into
considerable detail and illustrates several important aspects and properties of
option pricing.
A call option is a contract which grants the purchaser of the contract the
right, but not the obligation, to purchase a specic item called the underlying
asset from the seller of the contract at a specic date in the future called the
expiration date at a specic price called the strike price.
If the future price of the asset on the open market on the expiration date is
greater than the strike price, we say that the call option expires in the money.
In this case, the buyer exercises the option, buys the asset from the seller
at the strike price, sells the asset on the open market at the future price, and
makes a prot on the dierence.
If the future price of the asset on the open market on the expiration date is less
than the strike price, we say that the option expires out of the money. In this
case, the option is worthless, and the buyer loses his entire initial investment in
the option.
Our denitions above are for European options, which can only be exercised
on the expiration date. American options can be exercised on or at any time
before the expiration date. The math for European options is simpler, so we
use them here.
As an example which well use throughout this note, suppose an asset has a
current price of s = $100. There are two possible future prices: s1 = $120 with
probability p1 = 3/4 and s2 = $80 with probability p2 = 1/4. Consider a call
option on this asset with a strike price of x = $100.
Figure 1: Example
1 TWO-STATE OPTIONS 3
If the future price is $120, the option expires in the money, and it is worth
c1 = s1 x = $20. If the future price is $80, the option expires out of the
money, and it is worthless (c2 = 0).
What is the proper price c of this call option?
The expected value of the option is (3/4) 20 + (1/4) 0 = $15. This is not,
however, the proper price of the option. The proper price of the option turns
out to be $10. We give the standard proof of this from the literature.
The proof is by counter example, so we begin by assuming that you are willing
to buy this call option from me for $15. Im in fact eager to sell you an option
at this price, because I have a sure-re scheme to make a guaranteed prot of
$5 from the deal. My scheme involves a trick called hedging.
Heres how the scheme works:
In addition to selling you the option for $15, I also borrow $40 from the bank
or somewhere else. I now have $55. I pocket $5 and use the remaining $50
to buy 1/2 share of the asset. This 1/2 share in the underlying asset is my
hedge and is the key to making the scheme work. The fraction 1/2 is called the
hedge ratio, and well see how to compute it in section 2 where we work out
the simple algebra.
In our very simple two-state universe there are only two possible outcomes.
Either the asset increases in value to $120, or it decreases in value to $80.
First suppose that the ending price is $120. I sell my 1/2 share for $60. I use the
proceeds of this sale to repay my $40 loan. (For simplicity, we assume we dont
have to pay any interest on the loan. See section 2 for the details if we have to
pay interest.) Im left with $20. Because the ending price is greater than the
strike price, you exercise your option to buy one share from me for $100. You
pay me $100. I put that together with my $20, buy one share on the market for
the current price of $120, and give you that share. This total transaction comes
out even for me, and I still have the original $5 I pocketed at the beginning, so
my net prot is $5. Note that in this case you make $20 on your option, which
gives you a prot of $5 on your initial $15 investment in the option.
Now suppose that the ending price is $80. I sell my 1/2 share for $40 and use
that to repay my $40 loan. Your option has expired out of the money and so
you dont exercise it. Again, I still have the $5 I pocketed at the beginning,
so my net prot is $5. In this case you lose your entire $15 investment in the
option.
In either case, I still have the $5 I pocketed at the beginning. So I come out
ahead $5 in either case. I have taken advantage of the mispricing of the call
option to make a risk-free prot of $5.
Note that I do not have to come up with any of my own money to work this
scheme. Given a large enough number of other investors who are willing to buy
1 TWO-STATE OPTIONS 4
call options from me at a price of $15, and given a bank or someone else willing
to loan me large amounts of money, I can make an arbitrarily large amount of
money immediately with no risk at all, with a rate of return on my investment
of innity. This is the kind of opportunity investors dream about.
This kind of scheme to exploit mispriced assets to make a risk-free prot is
called arbitrage.
We can use the same arbitrage scheme to guarantee a risk-free prot for any
option price above $10. The dierence between the option price and $10 is the
guaranteed prot.
If the option is priced below $10, we can employ a similar arbitrage scheme
to guarantee a prot. In this case, we reverse everything and buy the option
instead of selling it, we lend money instead of borrowing it (e.g., we buy a
Treasury bill), and we sell a fraction of the asset short instead of buying it. The
details can be found in any good textbook on investing.
Thus, if the option is priced at any value other than $10, we can use arbitrage
to make a risk-free prot. So can anyone else who understands the trick and is
able to nd some sucker to sell the option to or buy the option from at the bad
price. Rest assured that every professional in the nancial world understands
the trick and is constantly on the lookout for suckers to exploit (its their job).
In the professional nancial world, if the option were priced at any value greater
than $10, everyone would want to sell the option and nobody would want to
buy it. Similarly, if the option were priced at any value less than $10, everyone
would want to buy it and nobody would want to sell it.
Thus, the options price must be exactly $10 in order for there to be any signif-
icant market in the option at all.
This completes our proof that the proper price for the option is $10.
At the beginning our proof we computed the expected value of the option to
be $15. The proper price turned out to be $10. Why is the expected value of
the option greater than its price? The explanation involves risk aversion. The
option is risky, and the extra $5 is a risk premium. If the expected value were
equal to the price, investing in the option would be a fair game, and risk-averse
investors would not be willing to purchase it. The $5 risk premium compensates
the investor for undertaking the risk of the investment.
As a brief digression, we must mention at this point that our example and our
analysis is ideal and ignores several important real-world issues. For example,
buying and selling fractional shares is often impossible or expensive. Also,
buying and selling assets and options on assets is not free as we have assumed.
Brokers, exchanges, advisors, dealers, and others all get fees to process these
kinds of transactions. We have ignored these fees. Also, our scheme involves
making and losing money in the nancial markets. These capital gains and losses
are taxable, and weve ignored the impact of these taxes. We have also made
1 TWO-STATE OPTIONS 5
the important assumption that the market for the underling asset is liquid,
which means that we can easily buy and sell shares of the asset at any time at
a well-dened current market price. In the real world dierent kinds of assets
have widely dierent liquidity. These kinds of pesky impediments to investing
are often called friction. Our ideal world is frictionless. The real world has
friction. To adjust our example for friction, we should say that the options price
must be exactly $10 or close to $10 in order for there to be any signicant
market in the option at all. Making this notion of close to more precise by
dealing formally with all the pesky frictions is a very dicult problem which we
are ignoring completely.
Now lets return to our arbitrage argument and look at it from another point
of view. Consider once again the strategy we used of rst borrowing $40 and
buying 1/2 share of the asset, then later selling the 1/2 share and repaying the
loan. This combination of borrowing and buying a fraction of a share is called
a synthetic call option.
In our arbitrage argument, we constructed our synthetic call option so that it
would have exactly the same payos in both possible outcomes as does the real
call option ($20 in one case, $0 in the other).
The law of one price states that in any situation in which two dierent invest-
ments have exactly the same payos in all possible future states of the world,
the investments must have the same price. The synthetic call option has a price
of (1/2) $100 $40 = $10. Thus the corresponding real option must also have
a price of $10.
Thus using the law of one price gives us another shorter proof that the proper
price for our option is $10.
Lets examine our example further to explore in a bit more detail these notions
of hedging, arbitrage and the law of one price.
In our scheme, we borrow money and buy a fraction of a share of the underlying
asset to oset the loss we might experience as a seller of the option. As a seller
of the option, we lose if the price rises, and we gain if the price falls. As a buyer
of the underlying asset, we gain if the price rises, and we lose if the price falls.
The two investments perfectly oset or hedge each other.
Borrowing money to help buy the fraction of a share is called leverage. In
our example, we buy 1/2 share for $50 and we borrow $40. Thus 4/5 of the
purchase price is borrowed. We say that our purchase is 80% leveraged.
What weve shown with our synthetic call is that in this particular example an
80% leveraged purchase of 1/2 share of the underlying asset is equivalent to one
call option on the asset with a strike price of $100. When we say equivalent
we mean that the two investments have exactly the same payos in all possible
future states of the world.
In our example, the price of the synthetic call is $10, and the corresponding real
1 TWO-STATE OPTIONS 6
option is priced at $15. Our arbitrage scheme boils down to simply selling the
overpriced investment (the real option) and buying the underpriced investment
(the synthetic call). Because the two investments hedge each other perfectly, we
run no risk, and we pocket the dierence in the prices, which is our $5 prot.
This is the essence of arbitrage. Whenever we nd a situation where the law
of one price is violated and two equivalent investments are selling at dierent
prices, we exploit the situation by selling the higher priced investment and
buying the lower priced one. The dierence in the prices is our prot. Because
the investments are equivalent, and because we bought one of them and sold
the other one, they hedge each other perfectly, and we make our prot with no
risk.
Signicant arbitrage opportunities are dicult to nd in the real nancial world.
The natural forces of the marketplace tend to eliminate them quickly when they
do appear. For example, suppose investments A and B are equivalent and A
is currently available on the market at a lower price than B. Arbitrageurs very
quickly notice the opportunity and rush to buy A and sell B. This drives up the
price of A and drives down the price of B until equilibrium is restored and the
arbitrage opportunity disappears. In todays very ecient nancial markets,
this process happens very quickly, often in only minutes.
There is no regulatory body that sets option prices to their proper price. This
is done by the natural forces of buyers and sellers in the marketplace. The law
of one price governs the marketplace. Arbitrageurs enforce the law. Punishment
for breaking the law is losing tons of money.
We now change the topic and return to our example to explore the role of the
probabilities of the possible outcomes. This is an important issue which we
havent covered yet.
Our arbitrage argument using our synthetic call has the interesting and impor-
tant property that it is independent of the probabilities of the possible outcomes.
The synthetic call constructed via a leveraged purchase of the underlying as-
set perfectly replicates the call option no matter what the probabilities of the
outcomes are.
Thus the price of the option in this universe with only two possible future states
is independent of the probabilities of those two states. Stated another way, in
this simple universe, the price of the option is independent of the expected rate
of return of the underlying asset. This seems strange at rst glance, but its
true.
The price does, however, depend on the volatility of the underlying asset. For
example, if our two possible ending prices are $140 and $80 instead of $120
and $80, the price of the call option is $13.33 instead of $10. In this case,
the replicating synthetic call is constructed by borrowing $53.33 and buying
2/3 shares of the asset for $66.67. See section 2 below for details on how to
calculate these values.
1 TWO-STATE OPTIONS 7
We see the same phenomenon in the Black-Scholes equation for much more
complicated real-life options for which the underlying assets have an innite
number of possible future prices instead of just two. Once again, the option
price given by the Black-Scholes equation is independent of the expected return
on the underlying asset, but depends signicantly on the volatility of the un-
derlying asset. As in the simple two-state universe, this is counter-intuitive but
nonetheless true. Well return to discuss this problem a bit more later.
For yet another point of view on the probabilities of the outcomes, lets return
to our original example, with the option mispriced at $15. We can think of my
selling the option to you at this price as a game we play. I use the arbitrage
trick, so the game boils down to the following essentials:
With probability 3/4, I win $5 and you win $5.
With probability 1/4, I win $5 and you lose $15.
Which side of this game would sane people rather play? The answer is obvious.
Now lets look at this same game with the proper option price of $10.
If we repeat the arbitrage scheme with the proper price of $10, my sure-re
prot disappears. If I do the synthetic call hedging trick, I win $0 in both cases.
You win $10 with probability 3/4 and lose $10 with probability 1/4. This game
is certainly more attractive to you as the option buyer, but its completely
unattractive to me as the option seller. Theres no point in my playing this
game with hedging, so I dont.
At the proper price of $10 my arbitrage scheme no longer generates any prot,
so I dont use it. Instead, however, I might simply sell you the option for $10 and
take my chances without hedging my bet by purchasing a fraction of a share
of the asset. Lets examine this possibility. This is called an unprotected
or naked call sale. For me, the outcome in this case is a loss of $10 with
probability 3/4 and a win of $10 with probability 1/4.
To summarize, at the proper price of $10 without the arbitrage scheme, the
game is the following:
With probability 3/4, I lose $10 and you win $10.
With probability 1/4, I win $10 and you lose $10.
In this game, the advantage is clearly with you, the buyer, not with me, the
seller. I would never play this game.
I might, however, play this game if I have a dierent opinion about the prob-
abilities of the outcomes. For example, if I believe that the $120 price has
probability 1/3 and the $80 price probability 2/3, I might be willing to sell you
an option at the price of $10.
To summarize, the probabilities of the outcomes do not enter into the determi-
nation of the option price. Investors subjective beliefs about the probabilities
1 TWO-STATE OPTIONS 8
do, however, determine who buys the options and who sells them.
1
What if the probabilities are so overwhelmingly in favor of the higher future
price possible outcome that everyone in their right mind considers the proper
price for the option to be very attractive? On the face of it, one might think
that this would tend to drive up the option price to balance buyers and sellers,
and that our argument must be wrong, because the proper price as set in
the marketplace for the option really does depend on the probabilities. This is
why option pricing in general and the Black-Scholes equation in particular are
so counter-intuitive.
Once again, we use our example to explore this further. Suppose, for example,
that nearly all knowledgeable analysts agree that the probability of the $120
ending price is 99/100, and the probability of the $80 ending price is 1/100.
Our option as priced at $10 is extremely attractive in this case, because we
have a 99% probability of ending up in the money and making a prot of $10.
We have only a 1% chance of losing our $10 investment in the option. Nearly
everyone should want to buy this option because its such a good deal. Shouldnt
this drive the price up?
This is indeed a problem. To get to the bottom of it, lets ignore the option for
a moment and instead focus on the underlying asset itself, which is currently
priced at $100. In our scenario, the consensus opinion is that theres a 99%
chance that the assets price will increase to $120, and only a 1% chance that
the assets price will fall to $80. This makes the asset itself a very attractive
investment, for exactly the same reason that the option on the asset is attractive.
Many more investors will want to buy the asset than will want to sell it at its
current price of $100. This will drive the assets price up. Note that as the asset
price rises, so does the price of the call option on the asset (see the equations
in section 2). Eventually equilibrium is reached, where both the asset price and
the option price reect the consensus opinion of investors concerning the future
prospects of the asset.
This resolves the conundrum. Yes, consensus probability beliefs do indeed af-
fect option prices, but only indirectly through the price of the underlying asset.
Stated another way, option pricing formulas for markets in equilibrium are inde-
pendent of the expected return on the underlying asset because this information
is already contained in the price of the underlying asset.
1
There are other important reasons for buying and selling options. For example, I might
sell you the option in our example because there is some other idiosyncratic characteristic of
my nancial life that exposes me to the risks of the underlying asset. My selling the option
to you in this case is a way for me to hedge that other risk I face, but that you do not
face. By exchanging risks in this way, we both benet from the transaction. This kind of
risk exchange is in fact the primary use of options in the nancial world, primarily by large
nancial institutions like banks.
2 THE ALGEBRA OF TWO-STATE CALL OPTIONS 9
2 The Algebra of Two-State Call Options
Note: In our example we made the simplifying assumption that the interest rate
for borrowing money is 0%. We did this to make the arithmetic easier for the
sake of exposition. In this section we include interest as a factor. Also, in our
example, the strike price happened to be the same as the current asset price.
This is also a non-critical assumption which we do not make in this section.
Let:
s = current asset price
c = call option price
s
1
= larger of the two possible future asset prices
s
2
= smaller of the two possible future asset prices
c
1
= call option value if future asset price is s
1
c
2
= call option value if future asset price is s
2
r = current yearly continuously compounded risk-free interest rate
t = time in years between the option sale and the option expiration
a = hedge ratio for synthetic call
b = amount borrowed for synthetic call
In the synthetic call, we buy a shares of the asset and borrow b dollars. The
price of this combination is as b. If the future asset price is s
1
, the payo is
as
1
be
rt
. If the future asset price is s
2
, the payo is as
2
be
rt
.
In the real call option, if the future asset price is s
1
, the payo is c
1
. If the
future asset price is s
2
, the payo is c
2
.
We want the synthetic call to replicate the real call. To accomplish this, we set
the payos in each state equal:
as
1
be
rt
= c
1
as
2
be
rt
= c
2
Solving these equations for a and b gives:
a =
c
1
c
2
s
1
s
2
b = (as
2
c
2
)e
rt
By the law of one price, the price of the real call option c must be the same as
the price of the synthetic call:
c = as b
2 THE ALGEBRA OF TWO-STATE CALL OPTIONS 10
In our example, we have s = 100, s
1
= 120, s
2
= 80, c
1
= 20, c
2
= 0, r = 0.
Plugging in these values gives:
a =
20 0
120 80
=
1
2
(hedge ratio)
b =

1
2
80 0

e
0t
= 40 (amount borrowed)
c =
1
2
100 40 = 10 (option price)
3 RISK-NEUTRAL VALUATION 11
3 Risk-Neutral Valuation
In section 1 we briey discussed how investors are risk-averse, and this is why
the value of an asset is less than its expected value.
We can, however, imagine an alternate universe in which investors are risk-
neutral instead of risk-averse. In this universe, investors dont care about risk,
they dont demand risk premia, and the value of an asset is simply its expected
value.
More properly speaking, because of the time value of money, in this universe
an assets value would be its future expected value discounted to present value
using the risk-free interest rate. To make things simpler, in this section were
going to pretend once again that our universe doesnt charge interest. As we
did in section 2, the reader is invited to rework the equations in this section to
accommodate interest. If you do this, you will nd that the main result of this
section still holds.
How would our trivial two-state options be priced in this universe? The arbi-
trage argument we used in our universe still works in the alternate risk-neutral
universe, so the equations in section 2 apply. Lets see what happens.
In the risk-neutral universe, the value of the underlying asset is its expected
value:
s = p
1
s
1
+ p
2
s
2
where p
2
= 1 p
1
Apply the equations we derived in section 2 for the value of the call option:
c = as b
=
c
1
c
2
s
1
s
2
s

c
1
c
2
s
1
s
2
s
2
c
2

=
c
1
c
2
s
1
s
2
[p
1
s
1
+ (1 p
1
)s
2
]
c
1
c
2
s
1
s
2
s
2
+ c
2
=
c
1
c
2
s
1
s
2
p
1
s
1
+
c
1
c
2
s
1
s
2
s
2

c
1
c
2
s
1
s
2
p
1
s
2

c
1
c
2
s
1
s
2
s
2
+ c
2
=
c
1
c
2
s
1
s
2
p1(s
1
s
2
) + c
2
= (c
1
c
2
)p
1
+ c
2
= p
1
c
1
+ (1 p
1
)c
2
= p
1
c
1
+ p
2
c
2
Thus, in the risk-neutral universe, the value of the call option is also its expected
value.
We can work out these same equations backwards. That is, if we start by
assuming that the value of the call option is its expected value, we can derive
our equations in section 2.
3 RISK-NEUTRAL VALUATION 12
This is an interesting result. In a risk-neutral universe, assets are priced quite
dierently than they are in our own risk-averse universe. In particular, in the
situation we are studying, both the underlying two-state asset and the two-state
call option on the asset are priced quite dierently in the two universes. What
weve seen, however, is that the equation in section 2 which tells us how to
compute the value of the option as a function of the value of the asset is exactly
the same in the two universes!
This is called the principle of risk-neutral valuation. It turns out that this
principle is also valid for real-life options where the underlying asset can have an
innite number of possible ending values. We examine this in detail in reference
[1].
REFERENCES 13
References
[1] John Norstad. Black-scholes the easy way.
http://www.norstad.org/nance, Feb 1999.